I understand there are many hardware devices (switches and routers) that traffic shape bandwidth based on SLAs to essentially to ensure response times. Just want to know when it should be employed.

1. In a shared environment? Are some customers willing to pay a few dollars more to get a basic amount of bandwidth?
2. Shared virtual environment? This one makes the most sense to me but how much do you charge for the min and how much for the bandwidth used over the ceiling rate (min - ceiling - max).
3. Dedicated environment? Why? To particular URLS? or Files?

Monthly bandwidth limits are common and customers pay for that. How about real-time bandwidth, are people willing to pay for it. And if so then how much?


Hopper

Learning the ropes before jumping in!

Hopper,
first, other than on ATM, all the QOS you can expect to see if what marketting genius have dubbed "soft QOS", which means "best effort".

And QOS is mostly used on a per-application basis. Even within a virtual private network, not everything needs high-priority, hence settings based on point-to-point definition, application type and time period.